Search

This paper presents a new supervised method for vessel segmentation in retinal images. This method remolds the task of segmentation as a problem of cross-modality data transformation from retinal image to vessel map. A wide and deep neural network with strong induction ability is proposed to model the transformation, and an efficient training strategy is presented. Instead of a single label of the center pixel, the network can output the label map of all pixels for a given image patch. Our approach outperforms reported state-of-the-art methods in terms of sensitivity, specificity and accuracy. The result of cross-training evaluation indicates its robustness to the training set. The approach needs no artificially designed feature and no preprocessing step, reducing the impact of subjective factors. The proposed method has the potential for application in image diagnosis of ophthalmologic diseases, and it may provide a new, general, high-performance computing framework for image segmentation.

Ulcer is one of the most common symptoms of many serious diseases in the human digestive tract. Especially for the ulcers in the small bowel where other procedures cannot adequately visualize, wireless capsule endoscopy (WCE) is increasingly being used in the diagnosis and clinical management. Because WCE generates large amount of images from the whole process of inspection, computer-aided detection of ulcer is considered an indispensable relief to clinicians. In this paper, a two-staged fully automated computer-aided detection system is proposed to detect ulcer from WCE images. In the first stage, we propose an effective saliency detection method based on multi-level superpixel representation to outline the ulcer candidates. To find the perceptually and semantically meaningful salient regions, we first segment the image into multi-level superpixel segmentations. Each level corresponds to different initial region sizes of the superpixels. Then we evaluate the corresponding saliency according to the color and texture features in superpixel region of each level. In the end, we fuse the saliency maps from all levels together to obtain the final saliency map. In the second stage, we apply the obtained saliency map to better encode the image features for the ulcer image recognition tasks. Because the ulcer mainly corresponds to the saliency region, we propose a saliency max-pooling method integrated with the Locality-constrained Linear Coding (LLC) method to characterize the images. Experiment results achieve promising 92.65% accuracy and 94.12% sensitivity, validating the effectiveness of the proposed method. Moreover, the comparison results show that our detection system outperforms the state-of-the-art methods on the ulcer classification task.

Cataloged from PDF version of article. The histopathological examination of tissue specimens is essential for cancer diagnosis and grading. However, this examination is subject to a considerable amount of observer variability as it mainly relies on visual interpretation of pathologists. To alleviate this problem, it is very important to develop computational quantitative tools, for which image segmentation constitutes the core step. In this paper, we introduce an effective and robust algorithm for the segmentation of histopathological tissue images. This algorithm incorporates the background knowledge of the tissue organization into segmentation. For this purpose, it quantifies spatial relations of cytological tissue components by constructing a graph and uses this graph to define new texture features for image segmentation. This new texture definition makes use of the idea of gray-level run-length matrices. However, it considers the runs of cytological components on a graph to form a matrix, instead of considering the runs of pixel intensities. Working with colon tissue images, our experiments demonstrate that the texture features extracted from "graph run-length matrices" lead to high segmentation accuracies, also providing a reasonable number of segmented regions. Compared with four other segmentation algorithms, the results show that the proposed algorithm is more effective in histopathological image segmentation

Currently, there is a lack of computational methods for the evaluation of mild traumatic brain injury (mTBI) from magnetic resonance imaging (MRI). Further, the development of automated analyses has been hindered by the subtle nature of mTBI abnormalities, which appear as low contrast MR regions. This paper proposes an approach that is able to detect mTBI lesions by combining both the high-level context and low-level visual information. The contextual model estimates the progression of the disease using subject information, such as the time since injury and the knowledge about the location of mTBI. The visual model utilizes texture features in MRI along with a probabilistic support vector machine to maximize the discrimination in unimodal MR images. These two models are fused to obtain a final estimate of the locations of the mTBI lesion. The models are tested using a novel rodent model of repeated mTBI dataset. The experimental results demonstrate that the fusion of both contextual and visual textural features outperforms other state-of-the-art approaches. Clinically, our approach has the potential to benefit both clinicians by speeding diagnosis and patients by improving clinical care.

Purpose: Pathologists often look at whole slide images (WSIs) at low magnification to find potentially important regions and then zoom in to higher magnification to perform more sophisticated analysis of the tissue structures. Many automated methods of WSI analysis attempt to preprocess the down-sampled image in order to select salient regions which are then further analyzed by a more computationally intensive step at full magnification. Although it can greatly reduce processing times, this process may lead to small potentially important regions being overlooked at low magnification. We propose a texture analysis technique to ease the processing of H&E stained WSIs by triaging clinically important regions. Method: Image patches randomly selected from the whole tissue area were divided into smaller tiles and Gaussian-like texture filters were applied to them. Texture filter responses from each tile were combined together and statistical measures were derived from their histograms of responses. Bag of visual words pipeline was then employed to combine extracted features from tiles to form one histogram of words per every image patch. A support vector machine classifier was trained using the calculated histograms of words to be able to distinguish between clinically relevant and irrelevant patches. Result: Experimental analysis on 5151 image patches from 10 patient cases (65 tissue slides) indicated that our proposed texture technique out-performed two previously proposed colour and intensity based methods with an area under the ROC curve of 0.87. Conclusion: Texture features can be employed to triage clinically important areas within large WSIs.

This paper presents a novel texture and shape priors based method for kidney segmentation in ultrasound (US) images. Texture features are extracted by applying a bank of Gabor filters on test images through a two-sided convolution strategy. The texture model is constructed via estimating the parameters of a set of mixtures of half-planed Gaussians using the expectation-maximization method. Through this texture model, the texture similarities of areas around the segmenting curve are measured in the inside and outside regions, respectively. We also present an iterative segmentation framework to combine the texture measures into the parametric shape model proposed by Leventon and Faugeras. Segmentation is implemented by calculating the parameters of the shape model to minimize a novel energy function. The goal of this energy function is to partition the test image into two regions, the inside one with high texture similarity and low texture variance, and the outside one with high texture variance. The effectiveness of this method is demonstrated through experimental results on both natural images and US data compared with other image segmentation methods and manual segmentation.

Vibro-acoustography is a recently developed imaging method based on the dynamic response of to low-frequency vibration produced by of ultrasound radiation force. The main differentiating feature of this method is that the image includes information about the dynamic properties of the object at the frequency of the vibration, which is normally much lower than the ultrasound frequency. Such information is not available from conventional ultrasound imaging. The purpose of this study is to evaluate the performance of vibro-acoustography in imaging mass lesions in soft tissue. Such lesions normally have elastic properties that are different from the surrounding tissue. Here, we first present a brief formulation of image formation in vibro-acoustography. Then we study vibro-acoustography of solid masses through computer simulation and in vitro experiments. Experiments are conducted on excised fixed liver tissues. Resulting images show lesions with enhanced boundary and often with distinctive textures relative to their background. The results suggest that vibro-acoustography maybe a clinically useful imaging modality for detection of mass lesions.

In many popular texture analysis methods, second or higher order statistics on the relation between pixel gray level values are stored in matrices. A high dimensional vector of predefined, nonadaptive features is then extracted from these matrices. Identifying a few consistently valuable features is important, as it improves classification reliability and enhances our understanding of the phenomena that we are modeling. Whatever sophisticated selection algorithm we use, there is a risk of selecting purely coincidental "good" feature sets, especially if we have a large number of features to choose from and the available data set is limited. In a unified approach to statistical texture feature extraction, we have used class distance and class difference matrices to obtain low dimensional adaptive feature vectors for texture classification. We have applied this approach to four relevant texture analysis methods. The new adaptive features outperformed the classical features when applied to the most difficult set of 45 Brodatz texture pairs. Class distance and difference matrices also clearly illustrated the difference in texture between cell nucleus images from two different prognostic classes of early ovarian cancer. For each of the texture analysis methods, one adaptive feature contained most of the discriminatory power of the method.

This paper describes the feasibility of selecting fractal feature vector based on M-band wavelet transform to classify ultrasonic liver images-normal liver, cirrhosis, and hepatoma. The proposed feature extraction algorithm is based on the spatial-frequency decomposition and fractal geometry. Various classification algorithms based on respective texture measurements and filter banks are presented and tested. Classifications for the three sets of ultrasonic liver images reveal that the fractal feature vector based on M-band wavelet transform is trustworthy. A hierarchical classifier, which is based on the proposed feature extraction algorithm is at least 96.7% accurate in the distinction between normal and abnormal liver images and is at least 93.6% accurate in the distinction between cirrhosis and hepatoma liver images. Additionally, the criterion for feature selection is specified and employed for performance comparisons herein.

This paper describes a new approach for texture characterization, based on nonseparable wavelet decomposition, and its application for the discrimination of visually similar diffuse diseases of liver. The proposed feature-extraction algorithm applies nonseparable quincunx wavelet transform and uses energies of the transformed regions to characterize textures. Classification experiments on a set of three different tissue types show that the scale/frequency approach, particularly one based on the nonseparable wavelet transform, could be a reliable method for a texture characterization and analysis of B-scan liver images. Comparison between the quincunx and the traditional wavelet decomposition suggests that the quincunx transform is more appropriate for characterization of noisy data, and practical applications, requiring description with lower rotational sensitivity.

Psoriasis is a chronic inflammatory skin disease that affects over 3% of the population. Various methods are currently used to evaluate psoriasis severity and to monitor therapeutic response. The PASI system of scoring is widely used for evaluating psoriasis severity. It employs a visual analogue scale to score the thickness, redness (erythema), and scaling of psoriasis lesions. However, PASI scores are subjective and suffer from poor inter- and intra-observer concordance. As an integral part of developing a reliable evaluation method for psoriasis, an algorithm is presented for segmenting scaling in 2-D digital images. The algorithm is believed to be the first to localize scaling directly in 2-D digital images. The scaling segmentation problem is treated as a classification and parameter estimation problem. A Markov random field (MRF) is used to smooth a pixel-wise classification from a support vector machine (SVM) that utilizes a feature space derived from image color and scaling texture. The training sets for the SVM are collected directly from the image being analyzed giving the algorithm more resilience to variations in lighting and skin type. The algorithm is shown to give reliable segmentation results when evaluated with images with different lighting conditions, skin types, and psoriasis types.

In this paper, we investigated the behavior of 22 co-occurrence statistics combined to six gray-scale quantization levels to classify breast lesions on ultrasound (BUS) images. The database of 436 BUS images used in this investigation was formed by 217 carcinoma and 219 benign lesions images. The region delimited by a minimum bounding rectangle around the lesion was employed to calculate the gray-level co-occurrence matrix (GLCM). Next, 22 co-occurrence statistics were computed regarding six quantization levels (8, 16, 32, 64, 128, and 256), four orientations (0° , 45° , 90° , and 135°), and ten distances (1, 2,...,10 pixels). Also, to reduce feature space dimensionality, texture descriptors of the same distance were averaged over all orientations, which is a common practice in the literature. Thereafter, the feature space was ranked using mutual information technique with minimal-redundancy-maximal-relevance (mRMR) criterion. Fisher linear discriminant analysis (FLDA) was applied to assess the discrimination power of texture features, by adding the first m-ranked features to the classification procedure iteratively until all of them were considered. The area under ROC curve (AUC) was used as figure of merit to measure the performance of the classifier. It was observed that averaging texture descriptors of a same distance impacts negatively the classification performance, since the best AUC of 0.81 was achieved with 32 gray levels and 109 features. On the other hand, regarding the single texture features (i.e., without averaging procedure), the quantization level does not impact the discrimination power, since AUC = 0.87 was obtained for the six quantization levels. Moreover, the number of features was reduced (between 17 and 24 features). The texture descriptors that contributed notably to distinguish breast lesions were contrast and correlation computed from GLCMs with orientation of 90° and distance more than five pixels.

Wireless capsule endoscopy is a revolutionary technology that allows physicians to examine the digestive tract of a human body in the minimum invasive way. Physicians can detect diseases such as blood-based abnormalities, polyps, ulcers, and Crohn's disease. Although this technology is really a marvel of our modern times, currently it suffers from two serious drawbacks: 1) frame rate is low (3 frames/s) and 2) no 3-D representation of the objects is captured from the camera of the capsule. In this paper we offer solutions (methodologies) that deal with each of the above issues improving the current technology without forcing hardware upgrades. These methodologies work synergistically to create smooth and visually friendly interpolated images from consecutive frames, while preserving the structure of the observed objects. They also extract and represent the texture of the surface of the digestive tract in 3-D. Thus the purpose of our methodology is not to reduce the time that the gastroenterologists need to spend to examine the video. On the contrary, the purpose is to enhance the video and therefore improve the viewing of the digestive tract leading to a more qualitative and efficient examination. The proposed work introduces 3-D capsule endoscopy textured results that have been welcomed by Digestive Specialists, Inc., Dayton, OH. Finally, illustrative results are given at the end of the paper.

In this paper, we propose the use of multiscale amplitude-modulation-frequency-modulation (AM-FM) methods for discriminating between normal and pathological retinal images. The method presented in this paper is tested using standard images from the early treatment diabetic retinopathy study. We use 120 regions of 40 × 40 pixels containing four types of lesions commonly associated with diabetic retinopathy (DR) and two types of normal retinal regions that were manually selected by a trained analyst. The region types included microaneurysms, exudates, neovascularization on the retina, hemorrhages, normal retinal background, and normal vessels patterns. The cumulative distribution functions of the instantaneous amplitude, the instantaneous frequency magnitude, and the relative instantaneous frequency angle from multiple scales are used as texture feature vectors. We use distance metrics between the extracted feature vectors to measure interstructure similarity. Our results demonstrate a statistical differentiation of normal retinal structures and pathological lesions based on AM-FM features. We further demonstrate our AM-FM methodology by applying it to classification of retinal images from the MESSIDOR database. Overall, the proposed methodology shows significant capability for use in automatic DR screening.

This paper focuses on extracting and analyzing different spectral features from transrectal ultrasound (TRUS) images for prostate cancer recognition. First, the information about the images' frequency domain features and spatial domain features are combined using a Gabor filter and then integrated with the expert radiologist's information to identify the highly suspicious regions of interest (ROIs). The next stage of the proposed algorithm is to scan each identified region in order to generate the corresponding 1-D signal that represents each region. For each ROI, possible spectral feature sets are constructed using different new geometrical features extracted from the power spectrum density (PSD) of each region's signal. Next, a classifier-based algorithm for feature selection using particle swarm optimization (PSO) is adopted and used to select the optimal feature subset from the constructed feature sets. A new spectral feature set for the TRUS images using estimation of signal parameters via rotational invariance technique (ESPRIT) is also constructed, and its ability to represent tissue texture is compared to the PSD-based spectral feature sets using the support vector machines (SVMs) classifier. The accuracy obtained ranges from 72.2% to 94.4%, with the best accuracy achieved by the ESPRIT feature set.

A fully automatic method is presented to detect abnormalities in frontal chest radiographs which are aggregated into an overall abnormality score. The method is aimed at finding abnormal signs of a diffuse textural nature, such as they are encountered in mass chest screening against tuberculosis (TB). The scheme starts with automatic segmentation of the lung fields, using active shape models. The segmentation is used to subdivide the lung fields into overlapping regions of various sizes. Texture features are extracted from each region, using the moments of responses to a multiscale filter bank. Additional "difference features" are obtained by subtracting feature vectors from corresponding regions in the left and right lung fields. A separate training set is constructed for each region. All regions are classified by voting among the k nearest neighbors, with leave-one-out. Next, the classification results of each region are combined, using a weighted multiplier in which regions with higher classification reliability weigh more heavily. This produces an abnormality score for each image. The method is evaluated on two databases. The first database was collected from a TB mass chest screening program, from which 147 images with textural abnormalities and 241 normal images were selected. Although this database contains many subtle abnormalities, the classification has a sensitivity of 0.86 at a specificity of 0.50 and an area under the receiver operating characteristic (ROC) curve of 0.820. The second database consist of 100 normal images and 100 abnormal images with interstitial disease. For this database, the results were a sensitivity of 0.97 at a specificity of 0.90 and an area under the ROC curve of 0.986.

Mass segmentation is used as the first step in many computer-aided diagnosis (CAD) systems for classification of breast masses as malignant or benign. The goal of this paper was to study the accuracy of an automated mass segmentation method developed in our laboratory, and to investigate the effect of the segmentation stage on the overall classification accuracy. The automated segmentation method was quantitatively compared with manual segmentation by two expert radiologists (R1 and R2) using three similarity or distance measures on a data set of 100 masses. The area overlap measures between R1 and R2, the computer and R1, and the computer and R2 were 0.76/spl plusmn/0.13,0.74 /spl plusmn/0.11, and 0.74/spl plusmn/0.13, respectively. The interobserver difference in these measures between the two radiologists was compared with the corresponding differences between the computer and the radiologists. Using three similarity measures and data from two radiologists, a total of six statistical tests were performed. The difference between the computer and the radiologist segmentation was significantly larger than the interobserver variability in only one test. Two sets of texture, morphological, and spiculation features, one based on the computer segmentation, and the other based on radiologist segmentation, were extracted from a data set of 249 films from 102 patients. A classifier based on stepwise feature selection and linear discriminant analysis was trained and tested using the two feature sets. The leave-one-case-out method was used for data sampling. For case-based classification, the area A/sub z/ under the receiver operating characteristic (ROC) curve was 0.89 and 0.88 for the feature sets based on the radiologist segmentation and computer segmentation, respectively. The difference between the two ROC curves was not statistically significant.